resending UP HOST Notifications

[[email protected]]

HI

We have a multiple Nagios system behind a firewalls that only allows messages out that we would like to monitor with a central Nagios Master.

Our intent was to configure the Nagios servers behind the firewall to send NSCA messages back to a centralized Nagios Master using the
notification functionality indicating the health of the host that it monitors.

Since it is behind a fiew wall, we would like that server to send the notification at regular intervals, so we can use it in a Passive service detection on the Central Master.

The problem is that once in an OK state, the resend of the notifications stops.
If it is in a failed state, the notification are sent at the retry_interval.

We need to have the resend of the OK / UP Notification at the retry_interval to continue in an OK / UP state.

Is there a way to configure this behaviour ?

Thanks

[slansing]

So what you are asking is how to configure Nagios to send Ok/Up notifications at the retry_interval that is set for checking when a service enters a non Ok/Up state? So you will always receive notifications of the current state even if it is Ok/Up?

[[email protected]]

Thanks for the reply .. and thanks for any help you can give me

Yes, but I think I should have said that they would be sent on the check_interval, not the retry_interval.

Notice in the following example that the CRITICAL is resent every 3 minutes where the OK is sent only once.

Jul 22 09:44:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 09:44:24, result is 2
Jul 22 09:47:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 09:47:24, result is 2
Jul 22 09:50:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 09:50:24, result is 2
Jul 22 09:53:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 09:53:24, result is 2
Jul 22 09:56:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 09:56:24, result is 2
Jul 22 09:59:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 09:59:24, result is 2
Jul 22 10:02:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;OK;notify-service-Master;nrpe ran this on 2013-07-22 10:02:24, result is 0
Jul 22 11:03:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 11:03:24, result is 2
Jul 22 11:06:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 11:06:24, result is 2
Jul 22 11:09:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 11:09:24, result is 2
Jul 22 11:12:28 my-fanag01 nagios: SERVICE NOTIFICATION: Master;my-fa-apu105;fa-apu105-updown;CRITICAL;notify-service-Master;nrpe ran this on 2013-07-22 11:12:24, result is 2

[slansing]

This may be possible, though I'm not quite sure what the reason would be, you would be overloaded with information. Nagios will notify you when a service goes into a warning/critical state, and once when it recovers.

[[email protected]]

Thanks for the reply

Maybe a better question would be ..

How do we configure a Nagios system that has multiple sites behind multiple firewalls all reporting to a centralized NOC.

Lets take one site for example.
Assume we have Nagios Server(B) and hosts(1-9) on one side of the firewall and another Nagios server(NOC) that is not.
(B) is suppose to report (NOC) so we can activly monitor the host(1-9).

Currently we are looking to solve the issue of servers behind a firewall disappearing. They cannot be check-host-alive or PINGed.

firewall
NOC  | B  (1-9)

Our initial design is to have the Nagios system behind the firewall report to a Nagios server(NOC) that is not behind the firewall using the notification logic and the NSCA plugin.

B is configured as PASSIVE host and (NOC) relies on the PROCESS_HOST_CHECK_RESULT and PROCESS_SERVICE_CHECK_RESULT which are configured to be sent by the NOTIFIER when a state changes.

On the (NOC) the HOST check_interval, retry_interval and freshness_threshold are set to expect a HOST_CHECK_RESULT every 10 minutes.
The check_command is set to check_dummy!0 (effective setting it to a OK state if a HOST_CHECK_RESULT is not fresh) since OKs are not resent.

If B crashes, it would act just like is was OK, sending nothing as the OK notification has aready been sent.

Hope this makes some sense

[lmiltchev]

Can you post the host/services configs (hide the sensitive data)?